Workplace e-learning for Automotive Assembly Operators · 2012-05-14 · in TPS [9]. The...

ICELW 2012 June 13th

-15th, New York, NY, USA

The International Conference on E-Learning in the Workplace 2012, www.icelw.org 1

Training Virtually Virtual Workplace e-learning for Automotive Assembly Operators

Lennart Malmsköld 1,2

, Roland Örtengren1 , Lars Svensson

3

1 Dep. of Product and Production Development, Chalmers University of Technology, Sweden 2 Volvo Group Trucks Technology, Gothenburg, Sweden

3 University West, Trollhattan Sweden

Abstract— This paper reports from a longitudinal study of a

Swedish manufacturer in the automotive industry, where a

series of studies have explored the potential and limitations

of computer-based training of assembly operators. The

study is focusing on two supplementing sets of target

variables – operators’ attitudes and the quality output from

operators in real production. Starting with small-scale

studies where proof-of-concept for virtual training is

established, the research moves on to comparative studies

where different computer-based learning models are

contrasted and evaluated. The research design ends with

large-scale field experiments assessing the effects of

computer-based training in terms of quality output. The

results clearly demonstrate that computer-based training,

when integrated with training of standardized work

procedures, outperforms traditional methods for operator

training, regardless of the content and the context of the

assembly operation. The findings of the study are

synthesized into a design framework for virtual training

where cognitive and craftsmanship training is contrasted to

the learning of product, process, sequence and finesse of

assembly.

Index Terms—Virtual Assembly Training, Automotive

Operator Training, Lean Production, Standardized Work.

I. INTRODUCTION

The rationale that drive initiatives of workplace e-learning ranges from the desire of reducing educational costs, through ambitions to create training environments where learners can operate without running the risk of causing serious harm to people or materials, to the necessity of bridging geographical, cultural or organizational distances. Similarly, the objectives of an e-learning initiative varies from the ambition to socialize newcomers into an existing work practice, to the need for preparing an organization to adapt to a future where new practices have to evolve [1].

In the automotive industry, the primary driving factor for launching e-learning initiatives stems from the fact that newly designed vehicles only exist in the virtual world right up to the point of the start of production. Car manufacturers have to strive for excellent quality performance, to be more flexible regarding production volumes and a high frequency of introducing new models. Furthermore, launches of a new vehicle has requirements where increased equipment reuse level, less prototype vehicles and decreased ramp-up time (the period from production starts until it has reached full production rate)

are characteristics. Usage of an existing production system means that introduction of new vehicles in the plant is done by mixing the new vehicles into the normal production flow during the launch period [2].

Earlier research shows some promising results for the potential of virtual training in manufacturing industry. Virtual reality (VR) systems has proven to be successful in areas such as welding [3], machining [4],[5] and object assembly [6],[7]. However, the positive results in prior research have yet to be proven in the context of full-scale realistic production.

II. THE AUTOMOTIVE CONTEXT

In automotive industries, the development process is based on the usage of math-based tools with the consequence that prototype material such as prototype vehicles are no longer produced at all or only in a limited number. To support the work with math-based tools in the development process, special virtual gates also exist in the00 development plan. Prior the gates, virtual builds are performed by Manufacturing Engineering, where complete vehicles are virtually assessed to verify that the design meets the requirements raised from a manufacturing point of view. In figure 1 some example from a virtual build is seen. Through this information is created which is possible to reuse for training purposes of the production organization in later stages.

Figure 1. Material created by the ME organization and used during a

virtual build




Training of assembly operators can differ in details but is in general based on the principle that a master teaches an apprentice. In the Toyota Production System, [8] which is the model for how many automotive companies are running their production the responsibility for the training and coaching of Team Members rests on the Team Leaders. Normally a Team Leader coaches a team containing 5-6 operators.

Standardized Work is another important concept included

in TPS [9]. The development of standardized work is

based on the use of standardized work documents. One of

them, the “Standard Operation Sheet (SOS)”, presented in

Figure 2, comprises a detailed description of the sequence

in which several operations (work elements) should be

performed. Each line in the table presented in the SOS is

representing a work element.

Another important document is the “Job Element Sheet”

(JES), illustrated in Figure 3. It describes each work

element shown in the SOS in detail by means of a list of

important work steps. Furthermore, any key points and

reasons connected to the operation are presented.

The training within TPS is named Job Instruction

Training (JIT) and the training has its origin in the TWI

(Training within Industry) concept [10]. The training is

divided into four steps.The first step concerns

familiarizing the Team Member with the work. The

second step is the demonstration of the operation, where

the JES and the components and tools employed in the

specific operation is used as training material. The third

step is a try-out, performed by the Team Member and

repeated until he/she masters the operation. The fourth

and final step of the training session is a follow and

feedback from the Team Leader

Training prior to and during pre-series production has

been the main target for this work and exemplified by the

conditions at the studied automotive company, is this

activity performed through a gradual integration of new

vehicles into existing production.

Figure 2. Example of a SOS from the studied automotive company

Figure 3. Example of a JES

In this setup, the vehicles in the first pre-series batch are

spread over time so that the interval between them is

often hours or initially even days. During succeeding pre-

series batches, the intervals gradually decrease. The Team

Leader can therefore perform all operations on the first

new vehicles by following them through all stations

within the team’s working area. By so, doing he/she also

demonstrates the operations to the Team Member at each

station. Ideally the JIT training, which is performed on a

separate vehicle off-line, should have started prior to the

first batch and historically this has generally been the

case. Due to cost and prototype reductions, the trend is

that JIT training before the start of pre-series production

is very limited or non-existent and is instead initiated

during or after the first pre-series batch.

III. METHOD

In view of the exploratory and empirical nature of the

problem domain, a case study approach appeared to be a

natural choice [11]. The major benefit of adopting a case

study approach is that the problem is studied in an

authentic setting with operators performing the learning

activities as an integrated part of their work. Primary data

of a quantitative character were studied in combination

with qualitative data. The latter approach was mainly

used in the first studies due to the relatively small groups

of test subjects. Therefore, semi-structured interviews

provided important complementary information for

understanding which factors and relations are relevant

when doing virtual training. The used analysis process for

the interviews was based on the method described by

[12].

In the two final studies the main conclusions were drawn

from data collected in existing quality follow up systems

and from performance data derived from the virtual

training tool. Semi-structured interviews and observations

were used to complement the collected data.

The methodological approach in the studies can be

described as data triangulation. This was applied in this

work to combine data from several sources such as

observations, interviews and quantitative data [13],[14].

In applied research the balance between rigour and

relevance must be considered in the research design [15].




The research presented in this work is the result of a

unique opportunity to perform studies in a real production

context, including scenarios during new vehicle model

launches where assembly was performed by experienced

operators. Thus, its relevance is very high. A high level of

rigour is ensured by the fact that the research design

incorporates test and control groups, a combination of

small and large scale studies in addition to a longitudinal

design.

IV. OVERVIEW OF THE STUDIES

In the present studies, the purpose was to measure and

understand the impact of virtual training, especially

during the pre-series period, prior new vehicle Start of

Production (SOP)

A schematic view and a timeline are presented in Figure

4. The studies in part A provided an understanding of the

impact of computer-based training compared to

traditional training, the results of which were documented

in [16]. A period of planning for the case studies then

followed. Interviews and tests took place, mainly with the

team of expert operators called Pilot Team. The output

from this period, Part B and the results from [16]

constituted the foundation for how the following studies

,in Part C, were performed. Input to [17] was also defined

in this intermediate period, since the first draft of the

framework (which constituted the major content of [17])

was formulated at this point. The framework presented in

[17] constituted the basis for the field studies reported in

[21] and [22].

V. CONTENT AND RESULTS OF THE STUDIES

In Part A the objective of the studies was to compare

virtual (computer-based) training with traditional

training. The study comprised operators divided into two

groups; a test group and a reference group. The test group

used the desktop based commercial VR tool, Vizendo,

see further description in [16]. The reference group

received traditional, instructor based training. All

activities took place prior to the start of pre-series

production and during the first period of a new vehicle

product launch. Assembly performance was compared

between the two groups during the first production on the

assembly line. A view of the virtual training and the

corresponding assembly line is shown in Figure 5.

Figure 4. An overview of the different parts of the work

Figure 5. Screenshot from training software and a photo from one of

the assembly stations

The results indicated that virtual training, combined with

motor skill training gained from the production of vehicle

batches in the pre-series phase, could provide operators

with similar assembly skills as those who receive

traditional training.

Part C of the research contained several studies and the

work in part C resulted in two papers ([18] and [17]).

A framework including knowledge phases in assembly

training was presented. The framework can be used for

understanding the learning processes in this specific area

but also for understanding operator training learning

trajectories as well as providing a theoretical foundation

for the design of novel techno-pedagogical virtual

training models [17].

In the process of learning new operations, two basic types

of learning can be identified: (a) Cognitive Learning and

(b) Associative and Autonomous Learning.

When studying the learning of a single operation or task,

cognitive learning is initially dominant, but after some

trials more focus is placed on performing the operation in

a correct way. At this stage the training of movement

patterns is dominant [19], [20]. The case studies reported

in [16] and [18] all demonstrated these learning types, but

since the work of an operator at an assembly station

alternates depending on which type of vehicle is to be

assembled, training of all combinations of operations is

necessary. In addition, quality requirements can influence

operations, and other aspects may have to be taken into

account to fulfil legal or customer requirements. The case

studies as well as the discussions and interviews with the

instructors in the assembly plant highlighted the different

phases of learning. Using this as a base, the framework

was derived with the four different knowledge phases of

Cognitive Learning and Associative and Autonomous

Learning. The knowledge phases can be seen as more or

less chronological steps in the learning process during

training. The framework and the knowledge phases are

shown in Figure 6. The framework can be seen as a

development of the work by [19] and [20] and constitutes

an elucidation model for this type of training as well as a

useful tool for framing the design space of virtual training

software for application within the specific area.




The studies that constituted the basis for [18] were carried

out based on the findings in [16], combined with input

from the expert operators (Pilot Team) gathered in Part B.

The focus was on cognitive learning processes and

mainly covered the first two cognitive dimensions of the

framework (the product and process phases). The

accomplished studies took place during the introduction

of new vehicle variants and constituted the basis for the

exploration of the two different models, Expert-based

Learning and System-based Learning (in the following

called the Expert Model and the System Model). Their

potential as a preparatory method for making cognitive

learning more effective was assessed.

The two learning models are similar in many respects.

Both use 3D data to describe new job content and are

intended to provide a complete visual presentation of

components to be assembled in the intended assembly

sequence. A central constituent of the Expert Model is an

active dialogue between different groups in order to

facilitate knowledge transfer (Figure 7).

The basic idea is to transfer knowledge from the experts

to the novices in three steps (Figure 7). However, as in

this situation the novice has experience of similar tasks

and specific conditions at the station where the operations

are performed, the activity can to some extent be seen as

mutual. In this model, the virtual tool, VisMockUp [18] is

used as visualization support for the Team Leader when

he/she describes the content of the new operations in a

dialogue with the Team Member. In addition, support

documents that describe each operation or job in detail

are employed.

In the System Model the new operational content is learnt

by self-study. The operators train by interacting with the

computer in a self-contained environment (Figure 8). In

the studies the tool used in the System Model was

Vizendo [18].

Figure 6. Knowledge phases and their connectedness in a design

framework.

Figure 7. Knowledge transfer in the Expert Model

Figure 8. Knowledge transfer in the System Model

The quantitaitve results [18] and the qualitative results from the interviews [18] indicated that the Expert Model was better than the System Model for preparatory training, since the dialogue between the Team Leader and Team Member supported by the viszualisation tool (VisMockUp) and process documents (JES/SOS) facilitated superior preparation for the new operations. This also was confirmed by the interviews [18].

In part D (Figure 4) the framework seen in Figure 6

constituted the base for the accomplished studies. The

two parts of cognitive learning; assembly sequence and

finesse, received special focus. The aim was to

understand whether or not computer-based training could

improve the operators’ learning and through this improve

their quality output. Several studies or field experiments

were conducted in two different areas of the assembly

line. The studies were carried out at a period when many

operations were relocated to new stations due to a change

in line speed and therefore training for new operational

sequences was required. In the studies, new developed

training software called SeQualia was used. SeQualia,

was created due to the need to enable operators to engage

in some form of self-study training using JIT information

as a complement to existing training. It was considered

important to include gaming features in the software, thus

incorporating an ‘edutainment’ dimension, i.e. combining

education and entertainment elements. The assembly

information, presented in a similar layout to the existing

process document, SOS and JES was retrieved from a

continuously updated JES/SOS database (Figure 9).




Figure 9. Main user interface in SeQualia, with a similar layout to a

SOS.

The major task for the operator is to create the correct

SOS for a selected station, model and variant. This is

accomplished through correct selection of one of four

randomly presented JES alternatives. If the operator’s

choice is incorrect, an error message appears on screen

and the red penalty bar at the bottom, which indicates the

accumulated number of incorrect answers in the training

session, increases by one unit. This procedure continues

until the whole SOS is completed.

As a means of tracking improvements related to

computer-based training, the two studies were set up as

field experiments [15] with a test and a reference group.

The test and reference groups belonged to different shifts

in the respective line areas and contained 4-5 teams, each

comprising one leader and five operators. In order to

address the research questions, quality performance was

compared between the reference operator groups that had

regular in-work training and test groups where some of

the in-work training was replaced by individual

computer-based training.

In both studies the subjects (Team Members) received an

individual introduction and demonstration of the

software. Each session lasted for approx. 30 minutes and

all Team Members selected one or several stations and

variants for training. The evaluation in both studies was

conducted by tracking quality losses through from the

existing quality data system and by means of interviews.

The first study, in part D, named Study A, was carried out

on a part of the main line, including 17 stations with 17

operators [21]. The operators in the test group had an

average total virtual training time of 2.6 hours each.

Interviews were performed with 6 operators and 3 Team

Leaders.

The second study, named Study B, took place in a sub-

assembly line area where doors were assembled. The line

area had 28 stations and all 28 operators were included in

the study. During the study period the operators had an

average total training time of 6.1 hours each.

Similar to Study A, data from the quality follow-up

system were tracked, but more specific analyses were

conducted based on the different types of quality

notation. At the end of the test period, interviews were

conducted with 6 operators and 2 Team Leaders.

The results of both studies revealed an overall

improvement in quality output for the test group

compared with the reference group. The results from

Study B were categorized into different fault types and in

several cases a direct connection between the computer-

based training process instructions (Key Points and

Reasons in the JES) and the quality loss levels was seen.

One of the quality problems studied in detail was leakage,

which is especially connected to doors and water

incorrectly drained from the door. Many Key Points and

Reasons were related to this type of problem and fault

type, which could therefore be connected to operator

knowledge of Key Points and Reasons. Figure 10

illustrates the frequency of the “Leakage” fault type for

both groups. The average improvement for the Test

Group between the Reference and the Study period was a

decrease from 6.5 down to 3.7 (-42%) and for the

Reference Group a decrease from an average of 4.6 to 4.4

(-5%).

The results of Studies A and B demonstrated an

improvement in quality output for the test groups

compared to the reference groups. Furthermore, the

operators described benefits of using the training

application as gaining a better understanding and

knowledge of the job content [21].

Part E, (Figure 4) contained several field experiments,

which evaluated how large-scale virtual operational

sequence training and related quality information can

support operators’ assembly performance.

All field experiments were conducted during the launch

of a new vehicle. In the first field experiment (Study ),

learning progress and quality performance were

compared between reference groups of operators whose

members only had regular training and a test group of

operators where some of the regular training was replaced

by individual computer-based training. The total number

of subjects was approximately 80 and included a test

group and two separate reference groups with 10-15

operators in each.

The experiment was conducted during one week of

intensive training where the test group had a session of

virtual training every day.

Figure 10. Loss levels regarding “Leakage” for both groups in study B




On the final day the virtual training tool (SeQualia) was

used to measure learning progress in the test and

reference groups respectively (Figure 11). In addition to

quantitative measures of operator learning progress due to

the virtual training, operator-related output quality during

the subsequent production weeks and questionnaires were

used to evaluate the effects of virtual training.

The results of the study (Figure 12) clearly demonstrated

that computer-based training outperforms traditional

training in terms of reading JES and SOS documents. The

mean value of average error per operator (AEO) after

four to five training sessions was substantially reduced. Since the reference group was relatively small, it was of

interest to obtain data that could be used to compare two

similar groups.

An additional reference group comprising 14 operators

selected from another line area was established, thus

permitting enhanced understanding of the impact of

virtual training [22].

In Study a large field experiment was conducted

covering the sub-plants from body-in-white to general

assembly. Focus was on validating the results of Study on a larger scale and approximately 360 operators each

received four twenty-minute training sessions over the

course of four training days, which occurred during a

production stop. The training activity focused on Key

Points and Reasons for each JES and all operational

content of the two stations in the operators’ own work

area was included in the training sessions. The SeQualia

training tool was used. Also Study Figureshowed

substantial learning progress: From around 6 errors in

initial mean values for each Plant Team (Average error

/operator, AEO), the levels dropped to below one error

per operator and session.

The quantitative results from the two field experiments

(andcombined with the positive attitudes expressed

by the operators and their Team Leaders indicate that this

is an effective way to train operators during new vehicle

launches in automotive production [22].

Figure 11. Training Setup for the different groups during the field

Figure 12. Operator performances in SeQualia, Study

Figure 13. Progress in theoretical skill for the four Plant Teams in Study

VI. A DUAL STEP CONCEPT FOR VIRTUAL TRAINING

– A CONTRIBUTION TO RESEARCH AND PRACTICE

Based on the case study results from all Parts, a Dual

Step Concept for virtual training is proposed and

constitutes the final contribution to research and practice

from this work. A cornerstone of the concept is the

framework presented in Figure 6. The concept is designed

to take account of the special conditions that exist in the

automotive industry, although the framework is more

general and can serve as a support in other research areas

connected to virtual training.

There are two cognitive knowledge phases of interest

when introducing new operations for skilled operators:

Product and Process aspects. Understanding of the job

content is the main requirement in preparatory training.

The combination of operations, the Assembly sequence, is

also of interest, but initially an understanding of operation

station location and general knowledge of the sequence is

sufficient. This content is the basis of the initial

preparatory training activity, Step 1. Here, 3D-models are

a necessary input, as no or very limited access to physical

components exists. In the next step intensive training of

complete sequences and finesse issues for the different




variants at each station are of more interest. This occurs

when the final sequence has been established, i.e. in the

later stages of pre-series period. Here the need for 3D

models for training purposes is limited, as the operators

already have good basic knowledge of the job content.

Furthermore, Figures 14 and 15 highlight the relevant

performance areas for virtual training and the proposed

timing of each step in the training activity.

Virtual training is an area where many concepts proposed in the literature have a basis in different types of 3D model-based self-study applications. This work explores several types of approach. The use of existing training methods, combined with the utilisation of 3D models was one of the approaches. Another was to reuse existing process documents such as SOS and JES for virtual training. The reason for exploring these alternative approaches was the interest in finding a successful virtual training concept, where reuse of existing information in true “lean spirit” is advantageous and a clear link to established training methods exists.

Figure 14. Progress in theoretical skill for the teams in Study

Figure 15. The timing of training steps 1 and 2 in the Dual Step virtual

training Concept

VII. CONCLUSIONS

This work has aimed at understanding if and how virtual training can serve as an effective alternative to traditional training and thereby support operators, especially when launching new vehicles.

The conclusions are:

i. A virtual training tool is an effective initial

training method and can replace traditional

instructor-based training in certain areas.

Using a desktop-based training tool with focus on

cognitive training was found to be effective for preparing

experienced operators before the introduction of new

vehicles.

ii. Virtual preparatory training of assembly

operators can be performed in an efficient

manner by using the Expert-based Learning

Model defined and explored.

The major reasons are:

The opportunity for dialogue between the Team

Leader and Team Member during initial

training,

Enhanced training for Team Leaders during the

training period

iii. Assembly quality performance can be

substantially improved through interactive

virtual training

The work has clearly demonstrated that

computer-based training outperforms traditional

training based on the reading of assembly

instructions (JES and SOS documents). Both the

theoretical learning rate and the quality

performance were improved by the training.

The four knowledge phases in the defined framework are

supported by the results of the case studies and have

formed the basis for the proposed Dual Step Concept for

virtual training.

The presented concept has the following parts:

A) Initial Virtual Preparatory training (Step 1)

Focus is on understanding of product and

process. This training, which is based on the

Expert Learning Model, is a preparation for the

first pre-production vehicles on the assembly

line. It serves as training for Team Leaders and

Expert Operators, who act as teachers during the

pre-production stage.




B) Craftsmanship Training – JIT

Initial practical training on training vehicles

using JIT and continued by gaining experience

of the production of pre-series vehicles on the

assembly line.

C) Virtual Sequence & Finesse Training (Step 2).

Here the focus is on drilling assembly sequences

and quality-related issues (finesse). This

training comprises interactive self-studies using

the System Learning Model.

VIII. REFERENCES

[1] L. Svensson “Inter-regional cooperation on Work-Integrated e-learning” , Journal of Workplace Learning, 16 (8), 2004

[2] M. Goyal, S. Netessine, T. Randall, ”Deployment of manufacturing flexibility: an empirical analysis of the North American automotive industry”, 2006, unpublished, [email protected].

[3] K. Fast, T. Gifford, R. Yancey, “Virtual Training for Welding”. Proceedings of the 3rd IEEE and ACM International Symposium on Mixed and Augmented Reality. Washington DC 2004, 298-299. ISBN 0-7695-2191-6

[4] A. Balijepalli, T. Kesavadas, “Value-Additon of Haptics in Operator Training for Complex Machining Tasks”, Journal of Computing and Information Science in Engineering, Volume 4, Issue 2, pp. 91-97, 2004

[5] A. Wasfy,T. Wasfy, A. Noor,”Intelligent virtual environment for process training”, Advances in Engineering Software, 35, pp: 337–355, 2004

[6] J. Dawei, A. Bhatti, S. Nahavandi,. “Design and evaluation of a haptically enable virtual environment for object assembly training”, Haptic Audio visual Environments and Games, 2009. HAVE 2009. IEEE International Workshop on, 7-8 Nov. 2009

[7] T. Gutiérrez, J. Rodríguez, Y. Vélaz, S. Casado, A. Suescu, and E. Sánchez, “IMA-VR: A Multimodal Virtual Training System for Skills Transfer in Industrial Maintenance and Assembly Tasks.” Conference proceedings, 19th IEEE International Symposium on Robot and Human Interactive Communication, 2010

[8] J. K. Liker, D. Meier, The Toyota Way Field book: A Practical Guide for Implementing Toyota’s 4Ps, 2006 New York: McGraw-Hill.

[9] D A. Dinero, Training Within Industry: The Foundation of Lean. Productivity Press, New York, NY, 2005

[10] P. Graupp, RJ. Wrona, THE TWI WORKBOOK: Essential Skills of Supervisors, Productivity Press ISBN:1563273152, 2006

[11] R. K. Yin, Case study research: design and methods 3. Ed. Thousand Oaks: Sage Publications, cop.,2003

[12] G. McCracken, The long interview: Newbury Park, Calif.: Sage, ISBN 0-8039-3352-5, 1988

[13] N.K. Denzin, The research act: A theoretical introduction to sociological methods. Chicago: Aldine, 1970.

[14] M.Q. Patton, Qualitative evaluation and research methods (2nd ed.). Newbury Park, CA: Sage, 1990

[15] P. Järvinen, On research methods, Tampere: Opinpaja, 1999

[16] L. Malmsköld, R. Örtengren, B. Carlson, P. Nylén, “Instructor based training versus computer based training - A comparative study”, Journal of Educational Technology Systems, 35 (4),2006-2007, pp. 457-478

[17] L. Malmsköld, R. Örtengren, B. Carlson, L. Svensson, “Virtual Training – towards a design framework”, G. Richards (Ed.), Proceedings of World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education 2007 (pp. 6299-6307). Chesapeake, VA: AACE.

[18] L. Malmsköld, R. Örtengren, B. Carlson, L. Svensson, “Preparatory virtual training of assembly operators – an explorative study of different learning models” Swedish Production Symposium (SPS) 2007, Conference proceedings

[19] P.M. Fitts, M.I. Posner Human performance, Belmont, Calif., Brooks/Cole Pub. Co. 1967 ISBN 0-13-445247-x

[20] E.M. Dar-El, K. Ayas, I. Gilad, “A dual-phase mode for the individual learning process in industrial tasks”. IIE Transactions, 27, 1995, pp 265–271.

[21] L. Malmsköld, R. Örtengren, B. Carlson, L. Svensson, “Computer Based Training for Automotive Assembly Operators – a Potential Way to Improve Quality Output” In conference proceedings, Interactive Technologies, 2008, Society for Applied Learning Technology

[22] L. Malmsköld, R. Örtengren, L. Svensson, “Improved quality output through computer based training - an automotive assembly field study”, unpublished

AUTHORS

L. Malmsköld, Dr was at SAAB Automobile Trollhattan, Sweden. He is now with Volvo Group Trucks Technologies, Process Development & Manufacturing Engineering, 405 08 Gothenburg, Sweden (e-mail: [email protected])

R. Örtengren, Professor, is with Department of Product and Production Development, Chalmers University of Technology, SE-412 96 Gothenburg (e-mail: [email protected])

L. Svensson, Ass. Professor is with Department Economics and IT, University West, SE-461 86 Trollhattan (e-mail: [email protected])

Workplace e-learning for Automotive Assembly Operators · 2012-05-14 · in TPS [9]. The...

Documents

Transcript of Workplace e-learning for Automotive Assembly Operators · 2012-05-14 · in TPS [9]. The...